Apple’s Mac OS X 10.4 brings productivity to desktops, turnkey power to servers, says InfoWorld.
Mac OS X Tiger burns bright
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
How many more Tiger reviews??? I could say Tiger has been reviewed to death!!!
So we have two contradictory articles in two days? I admit I didn’t RTFA as it needs registration, but how can you say that something that’s several orders of magnitude slower than Linux brings “turnkey power to servers”?
So we have two contradictory articles in two days? I admit I didn’t RTFA as it needs registration, but how can you say that something that’s several orders of magnitude slower than Linux brings “turnkey power to servers”?
————–
That was exactly my impression. Performance 8, scalability 9… What is this rating based on? Gut feeling?
Previously, Apple was often ignored by the mainstream IT press, except for iPods and the occasional iMac review. These articles show that Apple is getting media coverage, and some respect.
A truly excellent, well detailed article of what Is usually overlooked by other reviewers. I am keeping this article for re-reading because it is so good! Well done Tom. Just one criticism about the site. No print friendly option as far as I could see.
From the article:
“Darwin 8 will also build and boot on 32-bit x86 hardware. Yes, Darwin runs on x86…”
Hmmm… maybe Apple going to Intel hardware is the first step towards Apple as a software-only company after all. I didn’t think Steve had it in him.
>>Hmmm… maybe Apple going to Intel hardware is the first step towards Apple as a software-only company after all. I didn’t think Steve had it in him.
What has to do with being a only software company?
The current PowerMacs are PC98-PC2000 designs with OpenBoot ROMS. If Apple does switch to x86, then expect them to release a P4 PC2000 motherboard with OpenBoot ROMS (ie: no standard BIOS so Windows wont boot).
The Linux people will have Linux booting on the new system with-in 3 months of its release. Darwin is open source so you can see how the boot loader works; also, they currently boot Linux on PPC & OpenBoot so the changes should be minnor.
The *BSDs should be up within the same time frame.
MS Windows XP currently runs on PowerMac’s; MSFT just doesn’t release it widely (ie: you have to buy a developers kit for XBox2 to get a copy).
PS: The PC2000 standard is basically a standard PC without the ISA bus (used to support Floppy Disks, Serial, Printer, Keyboard, & Mouse on all current x86 mother boards {it was renamed to the SuperIO bus by Intel)). It also doesn’t have the standard interrupt controller (IRQ1-IRQ16); instead the PCI interrupt line is fed directly into the CPU.
So we have two contradictory articles in two days? I admit I didn’t RTFA as it needs registration, but how can you say that something that’s several orders of magnitude slower than Linux brings “turnkey power to servers”?
Well, for one, it was nowhere near several orders of magnitude. The author of this article also doesn’t bother to tell you what his standards for performance are, but based on other articles by him, it tends to be based on real world use where most servers don’t require the level of performance that benchmarks test. If your traffic is as high as the benchmarks can push things, you shouldn’t be using an XServe. For most purposes, though, an XServe is overkill – a G3 iMac could easily handle several thousand daily hits with PHP/MySQL content.
You mean that even though it MIGHT be slower, its still reliable and easy to use? And that for REAL WORLD use it works just as well? This is so disappointing, i think I’m going to have to recompile my front end and do some more pre-binding/ cache managing/ pref tweaking.
Is the Apple online store running on Mac OS X server?
iTunes Music store?
Apple.com is hosted with OS X Server… Not sure about the iTunes store, but I am sure it is also.
Not sure about the iTunes store, but I am sure it is also.
Yes. No ?
Hosted technically on OSX server but most of it’s handled by Akamai really, using Linux servers. In the same way, Microsoft.com runs on IIS but is mainly served from Linux machines
just on the off chance that someone missed it.
http://www.anandtech.com/mac/showdoc.aspx?i=2436&p=9
Wow! he can scroll pages of detailed search results in real time, something he wasn’t able to do in panther. I have a feeling that a switch to different hardware would be counter productive, it’s fairly obvious that they need to address plenty of OS issues before OSX can perform as well as linux on the same hardware.
“…it’s fairly obvious that they need to address plenty of OS issues before OSX can perform as well as linux on the same hardware”.
BS. How is it fairly obvious? Today we have two different articles contradicting each other. So, obvioulsy, it isn’t obvious.
Apple will be able to take a lot more open source for OS X86 and use better tools to develop OS X86.
Overall, whatever weaknesses are there today will be gone in 1-2 years after the move to X86.
OS X86 is the best news this year.
I’ll take Anand over infoweek anyday.
And I seriously doubt a switch to x86. Why would Apple ditch the chipmaker that finally made their machines competitive after such a short time, while they also stayed with a chipmaker that made their machines suck for many years?
I think the biggest thing that scares them is this; the possibility that in 6months time, we’ll see a Mac mini, using a PPC970FX 1.6Ghz running in these machines, possibly using a Intel graphics chipset – one part which Intel would like to grow, and their latest graphics chips sets are *VERY* competitive, and could easily beat the current 9200 included with the current Mac minis.
What Infoworld, and their IT rags hate, is Apple returning and growing. These companies make money off the missery induced by the x86 crap-o-la arhiceture, and the failure of Intel/AMD to bloody well kill of BIOS and and the numerous technologies that ‘promise’ plug and play.
IBM is climing back as an IT hardware powerhouse, but insted of worrying about the piddly, expensive side of the iT equation, namely, making and designing computers, they’re working with Apple and so forth. Its alot better for IBM in the long run to produce POWER processors to the tune of say 2millions per quarter than trying to piece together a desktop and workstation division that would have to be up against the likes of Dell.
What I would like to see in the next Mini-Mac is a faster hard disk, say a Hitachi 7200rpm, 1.6ghz PPC970FX processor, along with 512MB standard and a graphics card with atleast 128MB video memory. Per unit it may not make a heap of money, but if they soup up their software side of the house, expand iWork to include a spread sheet, database (say a slim down version of Filemaker), and buy out some more third party companies to expand the software portfoilio even more. Pull the customers in with the hardware, and push software at them like there is no tomorrow.
the one and only good review about OS X and XP ist still
http://www.xvsxp.com/
It’s not so biased and very detailed and have good arguments for any rating decision.
All the performance problems will undoubtedly be smoothed over once the transition to Intel processors is completed as http://news.com.com/Apple+to+ditch+IBM%2C+switch+to+Intel+chips… would suggest.
However, that does not rule out that there are glitches in the Tiger that I would like Apple to see addressed. I hope they’re not going to make me wait until the transition to the new architecture is completed.
Perhaps he meant “turkey power to servers”
Sorry, I couldn’t resist.(:-
… we miss the point. Apple has in actuallity far more standard and fixed archictures to support. Simply put they can focus on optimizing OSX pretty easily on these architectures. There are issues with the performance. Yes, there are. But OSX has been out there for just a few years. Take step back. See the big picture. In some ways (many ways) regarding usability OSX has reached where no other OS has. Judging from the pave that Apple penetrates the market and the way it developes OSX so far it’s just a matter of time before other issues (threading e.t.c.) are dealt with. Apple is a successful company with plenty talented, exceptional and committed software engineers – do not underestimate this fact. What we’ve seen till now is just a prelude of what these guys over at Apple can do.
Yes, but one article was well supported and credible and explained what computer scientists have said for 25 years: Micro kernels are slower than monoliths. More elegant, often, but always slower.
I’m sorry but real reviews mention pros and cons, are available free, and attempt to find real problems (everything has at least some real problem).
OS X is great for some things; but it’s not a performance OS. That doesn’t mean it can’t run as a high end server, it mean you’ll have to spend more on hardware to support it.
For anyone who doesnt want to sign up http://www.bugmenot.com .
Yes, but one article was well supported and credible and explained what computer scientists have said for 25 years: Micro kernels are slower than monoliths. More elegant, often, but always slower.
I’m sorry but real reviews mention pros and cons, are available free, and attempt to find real problems (everything has at least some real problem).
OS X is great for some things; but it’s not a performance OS. That doesn’t mean it can’t run as a high end server, it mean you’ll have to spend more on hardware to support it.
I would disagree with that assessement, its like stating that M:N threading is slower than 1:1 threading; YES, it is easy to squeeze performance out of a 1:1 threading because to implement it, is hardly a chor, HOWEVER, there are examples of over M:N threading have been implemented *VERY* well and *DO* outperform 1:1 threading, an example of this would be Tru64.
Regarding micro/mach kernels. like the above example, its easier to make a monolythic kernel, and it is fairly hard to make it majorly suck, however, on the other hand, micro/mach kernel will punish you big time for poor design decisions – the design and implementing IS complex, and to get performance to exceed monolythic can at times be difficulr, however, if one were to weigh up the complex nature of getting the performance in terms of development cost vs. the amount saved in the long run due to lower development costs – either with adding new features or maintaining the code base.
I’m sure if the Mach team were given more time, and more resources, the end result of their research could have been alot more than what see today. Only a fraction of the MacOS kernel is Mach, imagine of the WHOLE lot was Mach, not just the couple of parts right now – that would be an interesting thing to see; maybe thats what Apple should concertrate on; trying to ‘Mach-a-lise’ the whole kernel, by using the existing BSD code which they have.